A set of invariant amplitudes for the reaction $\ensuremath{\omega}\ensuremath{\pi}\ensuremath{\rightarrow}\ensuremath{\omega}\ensuremath{\pi}$ is constructed within the framework of the Veneziano model by expansions of the form $\ensuremath{\Sigma}(\mathrm{polynomials})\ifmmode\times\else\texttimes\fi{}(\mathrm{beta}\mathrm{function})$. The dominant asymptotic behavior corresponds to the $f$ trajectory in the $t$ channel and the $\ensuremath{\rho}$ and $B$ trajectories in the $s$ and $u$ channels. Separate treatments are given for the $\ensuremath{\rho}$ and $B$ contributions. In every case the leading trajectory poles have correct spin-parity as a consequence of enforcing proper asymptotic behavior on the invariant amplitudes. These relations are nontrivial and differ for the $\ensuremath{\rho}$ and $B$ contributions. The detailed construction of the amplitudes for $\ensuremath{\omega}\ensuremath{\pi}$ scattering differs from most previous applications because of the absence of an exotic channel. $\ensuremath{\omega}\ensuremath{\pi}$ scattering is notable in the number of terms required to represent the amplitude; a 72-parameter amplitude built using ${B}_{11}$, ${B}_{12}$, and ${B}_{21}$ cannot satisfy simple physical requirements for the $\ensuremath{\rho}$ trajectory. One must add further terms lacking the $\ensuremath{\rho}$ pole (such as ${B}_{22}$) in order to avoid decoupling the $\ensuremath{\rho}$ trajectory. The $B$ contribution is easily treated in direct analogy to the $\ensuremath{\rho}$ trajectory in $A\ensuremath{\pi}$ scattering. The implication of the hypothesis of partially conserved axial-vector current (PCAC) that the invariant amplitude ${T}_{1}$ (coefficient of ${e}^{\ensuremath{'}}\ifmmode\cdot\else\textperiodcentered\fi{}e$) should vanish at the Adler point is somewhat delicate because of the near degeneracy of the $\ensuremath{\omega}$ and $\ensuremath{\rho}$ mesons. The signature and amplitude conspiracy relations ensure the suppression of the apparent pole.
We consider a model in which the trace of the energy-momentum tensor, $\ensuremath{\Theta}\ensuremath{\equiv}{{\ensuremath{\Theta}}^{\ensuremath{\mu}}}_{\ensuremath{\mu}}$, contains both $\mathrm{SU}(3)\ifmmode\times\else\texttimes\fi{}\mathrm{SU}(3)$-invariant and -breaking terms. The latter are assumed to be proportional to ${S}_{0}$ and ${S}_{8}$, the $T=0$, ${J}^{P}={0}^{+}$ members of $(3,{3}^{*})\ensuremath{\bigoplus}({3}^{*},3)$. Each contribution to the trace is weighted by four minus the dimension of the associated local-field operator. Using Jacobi identity relations, we prove that the dimensions of ${S}_{0}$ and ${S}_{8}$ are equal. In order to investigate the consequences of the assumed chiral properties of the energy-momentum trace, we examine the $\ensuremath{\pi}\ensuremath{\pi}$, $\ensuremath{\pi}{A}_{1}$, and ${A}_{1}{A}_{1}$ matrix elements of $\ensuremath{\Theta}$. Ward-identity relations are employed to constrain the off-shell extensions of these matrix elements. In our model, the dimension $d$ of ${S}_{0}$ and ${S}_{8}$ is found to equal two in the $\mathrm{SU}(2)\ifmmode\times\else\texttimes\fi{}\mathrm{SU}(2)$ limit of massless pions. We also derive a more general relation for $d$ valid even when $\mathrm{SU}(2)\ifmmode\times\else\texttimes\fi{}\mathrm{SU}(2)$ is only an approximate symmetry. Scalar-dominance and smoothness assumptions enable us to extract additional information from this model. We predict $\frac{\ensuremath{\Gamma}({A}_{1}\ensuremath{\rightarrow}\ensuremath{\epsilon}\ensuremath{\pi})}{\ensuremath{\Gamma}(\ensuremath{\epsilon}\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi})}=0.07$ for the ${A}_{1}(1070)$ and $\ensuremath{\epsilon}(700)$ partial widths. Our model also leads to tentative conclusions concerning two subjects of current interest: the possibility that scalar mesons couple to $\ensuremath{\Theta}$ with a universal strength, and the existence of chiral-invariant contributions to $\ensuremath{\Theta}$.
Explicit minimal amplitudes are constructed for the reactions ${A}_{1}\ensuremath{\pi}\ensuremath{\rightarrow}{A}_{1}\ensuremath{\pi}$, ${A}_{1}\ensuremath{\pi}\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}$, and $\ensuremath{\pi}\ensuremath{\pi}\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}$. All amplitudes have correct Regge asymptotic behavior corresponding to the (degenerate) $\ensuremath{\rho}\ensuremath{-}f$ trajectory dominant in all channels. Signature is systematically enforced to leading order so that the poles on the leading trajectory have the correct spin and parity. The parity-conserving helicity amplitudes for ${A}_{1}\ensuremath{\pi}\ensuremath{\rightarrow}{A}_{1}\ensuremath{\pi}$ factorize, and multichannel factorization for the three coupled amplitudes is enforced for the leading trajectory. Multichannel factorization requires the extension of the ${A}_{1}\ensuremath{\pi}\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}$ amplitude, but the parameters in the new amplitude are all determined, except for two which can be identified with the (independent) ${A}_{1}\ensuremath{\rho}\ensuremath{\pi}$ couplings ${G}_{S}$ and ${G}_{D}$. Two acceptable ${A}_{1}\ensuremath{\pi}\ensuremath{\rightarrow}{A}_{1}\ensuremath{\pi}$ amplitudes are found, each involving two arbitrary constants in addition to ${f}_{\ensuremath{\rho}}={f}_{\ensuremath{\rho}\ensuremath{\pi}\ensuremath{\pi}}, {G}_{S}, {G}_{D}$, and the three ${A}_{1}{A}_{1}\ensuremath{\rho}$ couplings ${g}_{1}$, ${g}_{2}$, and ${g}_{3}$. These couplings are not constrained by the model except for one relation connecting ${G}_{D}$ and ${g}_{3}$. These results are compared with the hard-pion analysis by Schnitzer and Weinberg.
The scalar-dominance hypothesis, incorporating $\ensuremath{\epsilon}(700)\ensuremath{-}{\ensuremath{\epsilon}}^{\ensuremath{'}}(1060)$ mixing, is applied to a large class of elastic, single-particle matrix elements of the energy-momentum trace operator.